Clinical course and predictors in patients with borderline personality disorder during the COVID-19 outbreak: A 2.5-month naturalistic exploratory study in Spain

Dear Editor,

The COVID-19 outbreak is an emotional challenge for the general population. Within this context, people with prior mental disorders are at special risk of suffering additional and increasing problems (Yao et al., 2020). In particular, patients with borderline personality disorder (BPD) constitute a psychiatric cohort with poor personnel strategies to deal with stress and with high clinical support needs (Grambal et al., 2017). Thus, the COVID-19 outbreak represents a potential hazardous context that may negatively affect the clinical course of patients with BPD. Ultimately, this highlights the need to adopt new strategies to treat BPD patients during confinement beyond face-to-face therapies, including videoconferencing, phone calls or email messages. In Spain, the COVID-19 outbreak has been particularly dramatic, resulting in a period of restrictions from 14 March 2020 onwards, after the nationwide state of alert was activated.

The current exploratory, naturalistic study aimed to delineate the clinical course of patients with BPD during 2.5 months of home confinement and restrictions due to the COVID-19 outbreak and to address the effect of several clinical, psychosocial and treatment conditions throughout this period.

A total of 55 patients who met the DSM-5 (Diagnostic and Statistical Manual of Mental Disorders, 5th edition) criteria for BPD were recruited from the BPD program of the Adult Outpatient Mental Health Service of Mataró (Spain) at the beginning of home confinement on 14 March 2020. Prior to the COVID-19 outbreak, they took part in a naturalistic treatment consisting of weekly group psychotherapy (four groups of 12–15 patients each) and bimonthly individual treatment based on schema therapy. Prior BPD diagnosis and psychotherapeutic interventions were made by A.F., F.A., N.F. and C.P., who had at least 15 years of clinical experience in BPD. For the current study, the same authors also performed the telematic psychotherapy interventions and the subsequent assessments.

Change in clinical severity at the end of the 2.5-month period was the outcome variable. This was measured at the last individual phone session made during 25–29 May 2020, prior to restarting face-to-face psychotherapies, by using the Spanish version of the Clinical Global Impression – Improvement Scale (CGI-I; Pérez et al., 2007).

Due to the lack of prior literature, predictors were chosen intuitively and adapted to the limitations for more accurate data. Psychosocial predictors were assessed since the first phone session (16–20 March 2020) and included living status (alone or with someone) and loss or interruption of employment/academic course (not replaced by telematic procedures). Baseline clinical severity was assessed in the first phone session, prior to any type of phone psychotherapeutic intervention, and measured using the Spanish version of the CGI-I (Pérez et al., 2007). Treatment-related predictors during the 2.5-month period were number of phone psychotherapy sessions performed (see below) and other mental health interventions (e.g., outpatient care using phone calls by social workers and psychiatrists; attendance at psychiatric emergency services).

Due to the lack of accessibility for using videoconferencing in the healthcare service at the beginning of the COVID-19 outbreak, phone calls were chosen as the best approach to carry out psychotherapy. The proposed format extended over a 2.5-month period (from 16 March to 29 May 2020) and consisted of a maximum of seven individual sessions implemented every 9–11 days (range: 0–7). The first phone session took place during 16–20 March 2020 and the last during 25–29 May 2020. The mean duration of each session was 20 min (range: 10–30 min). At any session, if the patient did not pick up the phone at the first call, he/she was recalled twice throughout the same day. If possible, a message was left on the answerphone as a reminder for the following session.

Phone psychotherapy was cognitive behavioral-based. The main aim was to treat or prevent worsening of a primary BPD diagnosis during the aforementioned period. Techniques used included clarification, active listening, problem-solving skills, relaxation, debriefing, psychoeducation and distraction. The main issues addressed were isolation and interpersonal conflict, job loss, illness in a relative, daily habits and routine, and fear of contagion or negligence.

Because of the exploratory nature of the study, simultaneous multiple linear regression analysis was used to assess the predictors of change in clinical severity at the end of the 2.5-month period. Continuous predictors were baseline clinical severity and number of phone psychotherapy sessions. Categorical predictors were living status at home (0 = alone; 1 = with someone), job loss (0 = no; 1 = yes) and other types of mental health interventions (0 = no; 1 = yes). Statistical analyses were performed using the Statistical Package for the Social Sciences (SPSS, version 22.0).

A total of 50 patients (90.91%) completed both the baseline clinical severity assessment (first phone session) and the later clinical severity assessment (last phone session). The remaining five patients (9.09%) were not located at any time during the 2.5-month period and were excluded from further analyses.

Sociodemographic data for the 50 patients were as follows: 82% (n = 41) were female; mean age was 39.56 years (SD = 10.75); 22% (n = 11) lived alone; and 12% (n = 6) lost their job.

Patients completed an average of 6.62 phone psychotherapy sessions (SD = 0.69). During the 2.5-month period, patients needed additional interventions: phone call by social worker (n = 7), phone call by outpatient psychiatrist for medication adjustment (n = 8) and attendance at psychiatric emergency services (n = 5).

Baseline clinical severity was 5.48 (SD = 1.02), which is markedly ill. At the end of the 2.5-month period, patients did not report significant changes in clinical severity (mean = 3.80, SD = 0.67).

Simultaneous multiple linear regression analysis found that the predictors statistically accounted for the variability in clinical change at the end of the 2.5-month period (F _(5,44) = 4.91, p = 0.001, R² = 0.36). Living alone statistically predicted a poorer clinical course at the end of the period (β = −0.47, t = −3.68, p = 0.001). Other mental health interventions implemented were also associated with a poorer clinical course (β = 0.29, t = 2.18, p = 0.03). There were no other predictors that significantly contributed to the model, including the number of phone psychotherapy sessions (p > 0.05). There were no statistically significant correlations between predictors (p > 0.05).

The current naturalistic, exploratory study found that patients with BPD remained clinically unchanged throughout the 2.5-month assessment period, in which patients followed a substitutive phone psychotherapy intervention due to the restrictions caused by the COVID-19 outbreak.

Regarding the role of the proposed treatment, phone psychotherapy did not exert a dose-dependent effect on the clinical course. Due to ethical reasons, the lack of a control group with no psychotherapeutic treatment did not allow any confirmation of whether there was a qualitative therapeutic gain by receiving this substitutive treatment. In addition, treatments other than psychotherapy were associated with a poorer course over the referred period. This may be because those who worsened over time needed additional support that did not lead to clinical improvements at the time of the last assessment.

Interestingly, living alone was the most relevant predictor of a poorer clinical course. Due to the restrictions for physical contact with other people during the initial phases of the nationwide state of alert in Spain, it is hypothesized that BPD patients were not able to counterbalance the lack of relationships at home. Overall, social isolation would negatively affect their primary feelings of loneliness (Hauschild et al., 2018). It is likely that this finding was not specific for patients with BPD who lived alone but applied also to the general population in a similar home status during the COVID-19 outbreak. However, this would be more pronounced for patients with BPD due to their baseline clinical vulnerability. This finding highlights the need to address and improve psychosocial functioning in BPD patients, specifically those who are socially isolated.

The limitations of the current study are the lack of a comparison group, the need for more clinical assessments over the time period and the sample loss for the analyses.

Funding source

The corresponding author received funds for editing and proofreading the current manuscript from the Pla Estratègic de Recerca i Innovació en Salut of the Department of Health, Generalitat de Catalunya (SLT008/18/00175) .

Declaration of Competing Interest

None.